Kimi K2.6
Developed by Moonshot AI, Kimi K 2.6 is a native multimodal agentic model that excels in long-horizon coding, coding-driven design, and swarm-based task orchestration, with capabilities in transforming simple prompts into production-ready interfaces and workflows. Its Mixture-of-Experts (MoE) architecture, featuring 384 experts and a context window of 262,144 tokens, enables robust performance across various domains.
Input
Output
Context
262K
Max Output
262K
Parameters
1026.9B
Technical Specifications
Model TypeChat
Context Window262,144 tokens
Max Output Tokens262,144 tokens
Parameters1026.9B
Release DateApr 14, 2026
Training CutoffApr 1, 2025
Licenseother
Open SourceYes
Input Modalities
TextImage
Output Modalities
Text
Capabilities
Benchmarks
Artificial Analysis
35.9%
HLE
69.7%
AA-LCR
91.1%
GPQA
95.9%
Tau2-Bench
76.0%
IFBench
53.5%
SciCode
20.6%
Tau2 Banking
61.8
Coding
44.2
Intelligence
43.9%
Terminal-Bench Hard
65.9%
Terminal-Bench 2.1
More Models by Moonshot AI
Kimi Linear 48B A3B Base1000K context · 49.1B · Open weightsMoonlight 16B A3B Instruct8K context · 16B · Open weightsKimi K2 Instruct131K context · 1026.5B · Open weightsKimi K2.7 Code262K context · 1026.9B · Open weightsKimi K2 Thinking262K context · 1058.1B · Open weightsKimi Dev 72B131K context · 72.7B · Open weights
Related Tools
Data sourced from official provider APIs and documentation
Last updated: Aug 8, 2026
Your AI stack shouldn't stand still.
Every month new models become cheaper, faster, and more capable. Inferbase ensures your application automatically benefits without changing a single API call.